health-assessment

name	health-assessment
description	Assess customer health across product adoption, engagement, relationship, value realisation, and commercial dimensions. Identify at-risk accounts and recommend interventions.
argument-hint	[customer name, segment, or 'portfolio' for all accounts]
user-invocable	true
allowed-tools	Read, Write, Edit, Bash, Glob, Grep

Assess customer health for $ARGUMENTS.

Process (sequential — do not skip steps)

Step 1: Identify Data Sources

Before scoring, establish where health signals come from:

Signal type	Where to find it
Product usage	Analytics dashboard, event tracking, DAU/MAU metrics
Engagement	Login frequency, feature usage trends, session duration
Relationship	NPS/CSAT scores, meeting attendance, executive sponsor status
Value realisation	Customer-stated goals, ROI metrics, outcome tracking
Commercial	Billing status, contract terms, renewal date, pricing tier

If data sources are unavailable, note the gap — a health score with missing dimensions is less reliable but still useful.

Step 2: Score Each Dimension (0–100)

Score every dimension. Do not skip a dimension because data is sparse — estimate with documented reasoning.

Dimension	Weight	Scoring criteria
Product adoption	30%	Feature usage breadth, DAU/MAU ratio, time in app. 80+: using core + advanced features. 60–79: using core features. 40–59: sporadic use. <40: minimal or declining
Engagement	25%	Login frequency trend, support interactions, content consumption. 80+: increasing or stable high usage. 60–79: stable. 40–59: declining. <40: disengaged
Relationship	20%	NPS/CSAT score, executive sponsor engagement, meeting attendance. 80+: promoter, active sponsor. 60–79: passive, sponsor identified. 40–59: detractor or no sponsor. <40: unresponsive
Value realisation	15%	Achieving stated goals, ROI metrics, expansion signals. 80+: exceeding goals. 60–79: on track. 40–59: behind on goals. <40: no measurable value
Commercial	10%	Payment status, contract term remaining, pricing sensitivity. 80+: paid on time, long-term contract. 60–79: current, standard terms. 40–59: payment issues or pricing complaints. <40: at risk of non-renewal

Step 3: Calculate Composite Score

Composite = (Adoption × 0.30) + (Engagement × 0.25) + (Relationship × 0.20)
           + (Value × 0.15) + (Commercial × 0.10)

Step 4: Classify Health Status

Health	Score	Meaning	Response
Healthy	80–100	Customer is succeeding and engaged	Nurture. Identify expansion opportunities. Ask for referrals
Neutral	60–79	No immediate risk, but not thriving	Monitor. Quarterly check-in. Look for engagement opportunities
At Risk	40–59	Multiple warning signals present	Intervene within 1 week. Proactive outreach. Root cause analysis
Critical	0–39	Active churn risk	Escalate immediately. Intervention plan within 48 hours

Step 5: Identify Active Risk Signals

Check for churn risk indicators regardless of composite score — a single critical signal can override the composite:

Signal	Risk level	Response
Usage declining over 2+ weeks	Medium	Proactive check-in — "noticed you haven't used X recently"
Key feature not adopted after 30 days	High	Onboarding follow-up — offer training, remove friction
Support tickets increasing	Medium	Pattern analysis — is the product failing them?
Champion/sponsor left the company	High	Identify new sponsor immediately
NPS < 7 or CSAT declining	High	Personal outreach — understand the root cause
Payment issues (failed charge, downgrade inquiry)	Critical	Retention outreach same day
Competitor evaluation signals	Critical	Executive engagement + value reinforcement

Rule: A Healthy composite with a Critical signal = At Risk. The signal overrides the score.

Step 6: Recommend Interventions

For every account classified At Risk or Critical, define a specific intervention:

Intervention type	When to use	Actions
Engagement rescue	Usage declining, features unadopted	Training session, workflow review, success plan
Relationship repair	NPS low, sponsor gone, unresponsive	Executive outreach, new sponsor identification, business review
Value acceleration	Not achieving goals, no measurable ROI	Goal review, success metrics definition, use case workshop
Commercial save	Payment issues, downgrade inquiry, competitor evaluation	Pricing review, value demonstration, executive engagement

Each intervention has:

• Owner — who leads the intervention
• Timeline — when to start, when to review progress
• Success criteria — what does "saved" look like?
• Escalation — what happens if the intervention doesn't work?

Step 7: Portfolio Summary (if assessing multiple accounts)

When assessing more than one account, produce a consolidated portfolio view in addition to per-account detail. The portfolio output must include:

1. Per-account row table — one row per account showing: account name, per-dimension score (all 5), composite score, health classification, top risk signal, recommended intervention, owner.
2. Health distribution — count of accounts in each health tier (Healthy / Neutral / At Risk / Critical).
3. Prioritised at-risk list — Critical and At Risk accounts listed first, with the specific intervention required.
4. Cross-portfolio trends — which dimension is weakest across the at-risk cohort, common signals, systemic patterns. Don't just report per-account observations; surface what's true across the group.

Portfolio output template:

# Portfolio Health Assessment: [segment/portfolio name]

## Distribution
- Healthy: [n] | Neutral: [n] | At Risk: [n] | Critical: [n]

## Per-Account Scores

| Account | Adoption (30%) | Engagement (25%) | Relationship (20%) | Value (15%) | Commercial (10%) | Composite | Status | Top risk signal | Intervention | Owner |
|---|---|---|---|---|---|---|---|---|---|---|
| [name] | [0–100] | [0–100] | [0–100] | [0–100] | [0–100] | [weighted] | [tier] | [signal] | [action] | [person] |

## Prioritised Action List (Critical / At Risk first)

| Account | Status | Composite | Intervention | Timeline | Owner |
|---|---|---|---|---|---|
| [name] | [tier] | [score] | [action] | [start–review] | [person] |

## Portfolio Trends
- **Weakest dimension across at-risk cohort:** [dimension] — [observation]
- **Common signals:** [pattern]
- **Systemic recommendation:** [cross-account action, e.g. "onboarding gap — propose revised first-30-days programme"]

Anti-Patterns (NEVER do these)

• Scoring without data — estimate with documented reasoning, don't fabricate precision. "Adoption: 65 (estimated — no analytics dashboard available)" is honest
• Composite score only — always break down by dimension. A composite of 60 could be five dimensions at 60 or one at 100 and one at 0 — very different situations
• Ignoring single critical signals — a customer with a composite of 85 whose champion just left is At Risk, not Healthy
• Intervention without root cause — "schedule a call" is not an intervention plan. Why is the customer at risk? What specific problem will the intervention solve?
• One-time assessment — health monitoring is continuous. Define the re-assessment cadence (monthly for At Risk, quarterly for Neutral, semi-annually for Healthy)
• Expanding unhealthy accounts — expansion conversations with At Risk customers accelerate churn. Fix the health first

Output Format

# Customer Health Assessment: [account/segment/portfolio]

## Summary
- **Overall health:** [Healthy / Neutral / At Risk / Critical]
- **Composite score:** [0–100]
- **Assessment date:** [date]
- **Data confidence:** [High / Medium / Low — based on data availability]

## Dimension Scores

| Dimension | Weight | Score | Trend | Key signals |
|---|---|---|---|---|
| Product adoption | 30% | [0–100] | [↑ / → / ↓] | [specific evidence] |
| Engagement | 25% | [0–100] | [↑ / → / ↓] | [specific evidence] |
| Relationship | 20% | [0–100] | [↑ / → / ↓] | [specific evidence] |
| Value realisation | 15% | [0–100] | [↑ / → / ↓] | [specific evidence] |
| Commercial | 10% | [0–100] | [↑ / → / ↓] | [specific evidence] |

## Active Risk Signals
| Signal | Risk level | Detail |
|---|---|---|
| [signal] | [level] | [specific observation] |

## Recommended Interventions
| Intervention | Owner | Timeline | Success criteria |
|---|---|---|---|
| [action] | [person] | [start–review] | [what success looks like] |

## Next Assessment
- **Date:** [when]
- **Trigger for earlier review:** [what would warrant reassessing sooner]

Related Skills

• /customer-success:churn-analysis — when health assessment flags an account as at-risk, run a churn analysis to identify specific intervention opportunities.
• /customer-success:expansion-plan — when health assessment shows a healthy, engaged account, explore expansion opportunities.